Site icon Efficient Coder

Hermes Agent Complete Guide: 77 AI Skills and Slash Commands for Terminal Mastery

高效码农

Hermes Agent CLI Slash Commands Guide: Your Complete Handbook for AI-Powered Terminal Productivity

Hermes Agent is an AI-enhanced terminal tool built specifically for command-line workflows. It uses a clean slash-command system that lets you interact with your computer like a conversation: read and edit files, run shell commands, control browsers, and tap into dozens of pre-built skills. Whether you’re coding, researching, or creating content, these commands turn repetitive tasks into simple dialogue and complex multi-step projects into smooth, trackable flows.

This guide walks you through every core command, the skill system, task management, and real-world usage patterns. Everything is based directly on how Hermes Agent is designed, so you can start using it immediately and get real value from day one.

Core Tool Commands: The Everyday Foundation

Hermes Agent’s slash commands cover the basics first—files, terminal, browser, and code execution. They replace traditional tools with AI-friendly shortcuts that keep you in flow.

File Operation Commands

Most work starts with files. Hermes Agent gives you four straightforward commands:

  • /read – Reads the file and displays it with line numbers for easy reference.
  • /write – Overwrites the entire file with new content.
  • /search – Searches file contents or filenames. You can limit it to a specific path.
  • /patch – Finds and replaces text inside a file without opening an editor.

Example in practice:
/read ~/project/main.py
then
/write ~/notes.txt "Key notes here"
followed by
/search "TODO" --path ./src
and
/patch config.yaml "port: 3000" "port: 8080"

These commands feel natural because they match how you actually think about files—read, search, tweak, save.

Terminal Commands

When you still need raw shell access, use:

  • /term – Runs any shell command and returns the output.
  • /bg – Runs the command in the background so it doesn’t block your session.
  • /process – Manages background processes (list, poll status, or kill them).

You stay inside Hermes Agent without jumping between windows.

Browser Operation Commands

Hermes Agent can automate your browser:

  • /nav – Opens a webpage.
  • /click – Clicks a specific element on the page.
  • /type – Types text into a field.
  • /snap – Takes a snapshot of the current page.
  • /vision – Visually analyzes the page and answers your question.

Chain them together: navigate, click, type, then ask the vision tool to summarize. It turns manual browsing into programmable steps.

Code Execution Command

  • /exec – Runs Python scripts and can call Hermes Agent’s built-in tools.

This makes quick data tasks or automation scripts seamless.

Skill System Commands: 77 Pre-Built Skills That Do the Heavy Lifting

The real power of Hermes Agent comes from its 77 pre-loaded skills across development, research, creativity, and more. Skills are ready-to-use AI modules, not just scripts.

Managing Skills

  • /skills – Lists every available skill.
  • /skill – Shows details for one skill.
  • /skill_create – Creates a brand-new skill.
  • /skill_patch – Updates an existing skill.

Skills are grouped by category so you can find what you need quickly:

  • apple – macOS-only features (iMessage, reminders, notes).
  • autonomous-ai-agents – Orchestrates other AI agents like Claude Code or Codex.
  • creative – Generates ASCII art, charts, videos, and more.
  • data-science – Works with Jupyter kernels and data workflows.
  • github – Full GitHub operations (PRs, reviews, issues).
  • mlops – Machine-learning training, inference, and evaluation.
  • productivity – Connects to Google Workspace, Notion, Linear, etc.
  • research – Handles arXiv papers, blog monitoring, and academic tasks.
  • software-development – Includes TDD, debugging, and planning.

Check skills first, then dive in. If nothing fits, create your own with /skill_create.

Task Management, Memory, and Search Commands: Keep Everything Organized

Hermes Agent isn’t just for one-off commands. It helps you manage long-running work.

Task Management

  • /todo – Shows your current task list.
  • /todo_add – Adds a new task.
  • /todo_update – Updates progress (done, in-progress, blocked).

Memory and Session Search

  • /memory_add – Saves important information permanently.
  • /memory_remove – Deletes saved memory.
  • /search_sessions – Searches past conversations for similar problems.
  • /recent – Lists your latest sessions.

When you hit a familiar issue, /search_sessions pulls up the answer instantly. Save key facts with /memory_add and they stay available across sessions.

Agent Collaboration

  • /delegate – Hands a task off to a sub-agent for deep reasoning.
  • /delegate_batch – Runs multiple tasks in parallel.

It’s like giving Hermes Agent a team of helpers for heavy thinking.

Planning, Scheduling, Multimedia, and Utility Commands

Planning and Scheduling

  • /cron_create – Creates a recurring task.
  • /cron_list – Lists all scheduled tasks.
  • /cron_run – Runs a scheduled task right now.

Perfect for daily checks or automated reports.

Multimedia Commands

  • /tts – Converts text to speech.
  • /vision_analyze – Analyzes an image and answers your question.
  • /gif_search – Finds relevant GIFs.
  • /youtube – Gets the transcript of a YouTube video.

Utility Commands

  • /clarify – Asks you for clarification if something is unclear.
  • /plan – Switches to planning-only mode (no execution).
  • /debug – Starts systematic debugging.
  • /tdd – Enters test-driven development mode.

These let you adapt the tool to whatever mode your work needs.

Platform Integration Commands: Connect to the Tools You Already Use

Hermes Agent talks directly to external services:

GitHub Workflow

  • /gh_clone – Clones a repository.
  • /gh_pr

Exit mobile version